Set Up Agentforce for Analytics

Assign User Permission Sets

1. From Setup, in the Quick Find box, enter Permission, and select Permission Sets.
2. Select a permission set: Tableau Unmetered Admin or Tableau Next Admin, Tableau Unmetered Platform Analyst or Tableau Next Platform Analyst, Tableau Unmetered Self-Service Analyst or Tableau Next Self-Service Analyst, Tableau Unmetered Self-Service Consumer or Tableau Next Self-Service Consumer, or Tableau Next Consumer (Unmetered) or Tableau Next Consumer.
3. Click Manage Assignments.
4. Click Add Assignments.
5. Select one or more users by checking the box next to their name.
6. Click Assign.
7. Perform these steps again for the same users, selecting the Data Cloud User permission set.

Enable Tableau Next, Einstein, and Agentforce

Agentforce for Analytics requires that Tableau Next, Einstein, and Agentforce are set up and configured in your organization. Perform these steps if you haven't yet.

1. Enable Tableau Next.
2. Enable Agentforce.
3. Enable Einstein Generative AI for Tableau Next.
   1. Set up Einstein Generative AI.
   2. Set up Einstein Trust Layer.
   3. Turn on Einstein Generative AI data collection and storage.

Activate Agentforce for Analytics

Agentforce for Analytics is a collection of pre-built agentic capabilities made available through the Analytics and Visualization agent template. Turning these settings on in Tableau Next makes the Analytics and Visualization agent template and its topics and actions available to you in Agentforce Studio.

1. In Tableau Next, click Administration, and select Settings.
2. Turn on Agentforce for Analytics.
3. Turn on agent capabilities.

   Concierge: Analytics Q&A	Includes the Data Analysis subagent.
   Data Pro - Semantic Model Creation	Includes the Data Pro - Preparation and Modeling subagent.
   Inspector: Proactive Alerts	Includes the Data Alert Management subagent.

4. Optional: Turn on semantic layer AI-readiness features. These settings aren't required for Agentforce for Analytics, but they enable useful tools for improving AI readiness in semantic models used by analytics agents.

   Semantic Built-In AI	This accelerates and streamlines semantic model creation and maintenance.
   Semantic Model AI Optimization (Beta)	This enables automated checks on semantic data models to detect optimization opportunities. This tool surfaces issues with guided resolutions to improve model quality and reliability for AI-powered agents and analytics.
   Semantic Description Generator (Beta)	This drafts descriptions for your semantic model entities with AI. Accurate descriptions boost agent response accuracy. Be sure to review and edit suggestions before saving.
   Q&A Calibration (Beta)	You can access this tool in the Tableau Next menu to test questions in Concierge, and then verify and calibrate results.

Note

Analytics agents require a well-designed semantic model to understand data and confidently answer questions. AI-readiness features add context, enrich, and calibrate semantic models so you can actively test, refine, and deepen an agent's understanding of your business.

Create an Agent with the Analytics and Visualization Template

Use the Analytics and Visualization template to create a new agent and assign it to users.

1. Create an Agent from an Agentforce Employee Agent Template using these settings:
   1. Select the Analytics and Visualization template.
   2. Add subagents.

   Data Analysis	Includes the Analyze Data action.
   Data Pro - Preparation and Modeling	Includes the Create Calculated Field and Create Semantic Data Model actions.
   Data Alert Management	Includes the Create Data Alert action.

2. To use an agent, users need a permission set that grants them access. Manage Employee Agent Access for the users who need access to the new agent.

Verify the New Agent in Tableau Next

Users can Select Agent and choose the new agent from Agentforce in Tableau Next.

1. Open a metric or dashboard in Tableau Next.
2. Open Agentforce.
3. In the Select Agent list, choose the new agent.
4. Ask a question about the asset and verify a response is generated.